Data centers have become a core component of modern living, by containing and distributing the information required to participate in everything from social life to economy. In 2017, data centers consumed 3 percent of the world’s electricity, and new technologies are only increasing their energy demand. The growth of high-performance computing — as well as answers to growing cyber-security threats and efficiency concerns — are dictating the development of the next generation of data centers.

But what will these new data centers need in order to overcome the challenges the industry faces? Here is a look at 5 major trends that will impact data center design in the future.

1. Hyperscale functionality

The largest companies in the world are increasingly consolidating computing power in massive, highly efficient hyperscale data centers that can keep up with the increasing demands of enterprise applications. These powerful data centers are mostly owned by tech giants like Amazon or Facebook, and there are currently around 490 of them in existence with more than 100 more in development. It’s estimated that these behemoths will contain more than 50 percent of all data that passes through data centers by 2021, as companies take advantage of their immense capabilities to implement modern business intelligence solutions and grapple with the computing requirements of the Internet of Things (IoT).

2. Liquid efficiency

The efficiency of data centers is both an environmental concern and a large-scale economic issue for operators. Enterprises in diverse industries from automotive design to financial forecasting are implementing and relying on machine-learning in their applications, which results in more expensive and high-temperature data center infrastructure. It’s widely known that power and cooling represent the biggest costs that data center owners have to contend with, but new technologies are emerging to combat this threat. Liquid cooling is swiftly becoming more popular for those building new data centers, because of its incredible efficiency and its ability to future-proof data centers against the increasing heat being generated by demand for high-performance computing. The market is expected to grow to $2.5 billion by 2025 as a result.

3. AI monitoring

Monitoring software that implements the critical advances made in machine learning and artificial intelligence is one of the most successful technologies that data center operators have put into practice to improve efficiency. Machines are much more capable of reading and predicting the needs of data centers second to second than their human counterparts, and with their assistance operators can manipulate cooling solutions and power usage in order to dramatically increase energy efficiency.

4. DNA storage

In the two-year span between 2015 and 2017, more data was created than in all of preceding history. As this exponential growth continues, we may soon see the sheer quantity of data outstrip the ability of hard drives to capture it. But researchers are exploring the possibility of storing this immense amount of data within DNA, as it is said that a single gram of DNA is capable of storing 215 million gigabytes of information. DNA storage could provide a viable solution to the limitations of encoding on silicon storage devices, and meet the requirements of an ever-increasing number of data centers despite land constraints near urban areas. But it comes with its own drawbacks. Although it has improved considerably, it is still expensive and extremely slow to write data to DNA. Furthermore, getting data back from DNA involves sequencing it, and decoding files and finding / retrieving specific files stored on DNA is a major challenge. However, according to Microsoft research data, algorithms currently being developed may lower the cost of sequencing and synthesizing DNA plunge to levels that make it feasible in the future.

5. Dynamic security

The average cost of a cyber-attack to the impacted businesses will be more than $150 million by 2020, and data centers are at the center of the modern data security fight. Colocation facilities have to contend with the security protocols of multiple customers, and the march of data into the cloud means that hackers can gain access to it through multiple devices or applications. New physical and cloud security features are going to be critical for the evolution of the data center industry, including biometric security measures on-site to prevent physical access by even the most committed thieves or hackers. More strict security guidelines for cloud applications and on-site data storage will be a major competitive advantage for the most effective data center operators going forward as cyber-attacks grow more costly and more frequent. The digital economy is growing more dense and complex every single day, and data center builders and operators need to upgrade and build with the rising demand for artificial intelligence and machine learning in mind. This will make it necessary for greener, more automated, more efficient and more secure data centers to be able to safely host the services of the next generation of digital companies.

Author: Gavin Flynn

Source: Information-management

Every year, the technology industry seems to come up with new products that have the capability to manage themselves. From cars that tell us if we’re backing up too fast to AC units that turn on when they realize the residents are on their way home, we’re seeing technology continuing to advance in their ability to self-manage.

The next logical step we are seeing is self-managing data centers, where automation and machine learning handle administrative storage tasks.

Even for those who don’t believe machines can execute the tasks of an IT manager more effectively than their human counterparts, the efficiency gains from offloading repetitive functions -- or making connections between dissimilar, often unrecognized events – should give businesses the ability focus on strategic objectives that will help the company flourish.

Like self-driving cars, the self-managed data center that rarely needs human intervention could be coming sooner than we think. Data centers are increasingly utilizing full self-managed capabilities, which wouldn’t be possible without automation and machine learning technology. 

Below are the three main trends that are helping to make self-managed data centers a reality.

Promising performance without intervention

Automation and machine learning offer multiple capabilities that aid in developing the self-managed data center. 

One is that organizations can guarantee performance without intervention. With traditional storage, applications compete for resources from a fixed number of buckets or IOPS. Guaranteeing a set number of IOPS for a particular application prevents organizations from accessing those IOPS for other apps. 

Automation enables organizations to access IOPS resources and allows virtual machines (VMs) to employ them for other necessary purposes. So, although it ensures a clear lane for every VM, it also enables the VMs to access IOPS as necessary. 

This approach avoids the danger of saving and wasting unused IOPS, instead making them available when needed.

Ensuring a clear lane for very every virtual machine

In the future, machine learning and automation promise to optimize the performance of storage arrays and predict future usage trends. It can analyze past performance to predict trends for the next two months, for example, giving organizations insight into what’s necessary to optimize performance and capacity for storage-array pools.

Machine learning should enable organizations to move VMs from a particular array to somewhere else in the pool if through its ability to analyze performance trends. Furthermore, it would allow organizations to predict and address poor performance on an array.

Machine learning can also help businesses plan for their future. Analytics would enable organizations to improve predictions and make savvier decisions about infrastructure requirements to avoid downtime. It’s like building another wing on an apartment complex to address growing resident occupancy in the future.

Optimizing the performance of storage arrays and predicting future usage trends

Additionally, by giving each VM its own lane, organizations could make optimal use of all their performance all the time. On those rare occasions when VMs ask for more than the storage can deliver, performance could be assigned dynamically to applications that require it rather than on a first-in, first-out basis.

When further development of apps and devices that use machine learning take place, companies will try to find new and exciting ways to incorporate AI.

The controversial debates about AI will continue, but there are ways to utilize it without going overboard and giving over too much control. 

Automated, self-managed data centers are becoming a reality, promising real-time, predictable performance without IT intervention. Even dense IT infrastructure that’s typically difficult and time-consuming to upgrade and control is becoming automated and divided into elements managed through software instead of hardware. These data centers are increasingly utilizing full self-managing capabilities. 

Ultimately, with the combination of AI and machine learning, IT teams should finally have the ability to focus their time on more important tasks that add real value to the company rather than being stuck in the back end of the data center. 

The data center that manages itself and rarely needs assistance has the potential to arrive sooner than expected. In the coming months, you’ll start to see how machine-learning-based intelligent automation will become a critical component of the modern data centers

Author: Chris Colotti

Source: Information Management